Good backtest result.

To add comments, please log in or register
texto
32
texto  

Hi,


I wanted to know what is a good backtest result in terms of profit. For e.g if I backtest an EA that makes 4000 profit in 4 years with a starting capital 5000 - would that be considered acceptable? Equity curve rises steadily and the absolute drawdown is 150.

How do you guage your backtest results?


Thanks.

Simon Gniadkowski
17239
Simon Gniadkowski  
How good is the data you are using for your Backtest ? is the spread realistic ?
seravitae
94
seravitae  

'Acceptable' is a loose term. Backtests perform analysis on market data that is already known. In the future, your EA could totally fail depending on your EA's rules. You can get more confidence with a forward test, but these are also not bulletproof in terms of getting certainty.

In terms of how 'acceptable' an EA might be financially, comes from a lot of variables. Acceptable might be varied when considering total risk, ratio of wins to losses, average win $$, average loss $$, average trade time, average balance/profit growth.. The list goes on... Acceptable in the easiest terms is "do you feel happy to use it". There will always be EAs out there that make more and less money than your EA, so don't benchmark your EA against anyone elses.

Simon Gniadkowski
17239
Simon Gniadkowski  
seravitae:

'Acceptable' is a loose term. Backtests perform analysis on market data that is already known. In the future, your EA could totally fail depending on your EA's rules. You can get more confidence with a forward test, but these are also not bulletproof in terms of getting certainty.

A forward test is just a very slow backtest . .
To add comments, please log in or register